Friday, July 29, 2011
A Plan for Forking Wikipedia to Provide a Reliable Secondary Source on Law
Recently Wikipedia rolled out a feedback feature (example at right) that allows readers to rate the page they are looking at. You can give the page a score from one to five in each of four categories: Trustworthy, Objective, Complete, and Well-written. Then, there's an optional box you can check that says "I am highly knowledgeable about this topic."
This may be good for flagging pages that need work. But, of course, Wikipedia's trustworthiness problem is not going to be solved by anonymous users, self-declared to be "highly knowledgeable," deeming articles to be trustworthy.
I do think, however, that if you forked Wikipedia to create a version with an authenticated expert editorship, then the ratings could evolve Wikipedia content into being a credible source. In fact, I think it could work well for articles on law.
"Forking," for software developers, means taking an open-source project and spliting off a version that is then developed separately. The open-source license specifically allows you to do this. Forking a project is not always productive, but I think it could be useful in creating an encyclopedia-type reference about the law that is both trustworthy and freely available.
There's a real need for such material in the legal sphere. Right now, there seems to be an accessibility/credibility trade off with secondary sources of legal information: Wikipedia is accessible, but not credible. Traditional binder sets are credible, but not very accessible – using them is generally either expensive (in terms of subscriptions fees) or burdensome (via a trip to the nearest law library).
If, however, you could take Wikipedia and apply credibility on top of it, you would have a secondary source that is both credible and accessible.
Imagine grabbing all the Wikipedia pages about law – which at this point are generally very well developed – and then, while continuing to make them viewable to the public, locking them so that only authenticated lawyers and law students could edit them. These expert editors could then correct errors where they find them. Where they don't find errors, they could click a box indicating trustworthiness. As time went on, pages would have errors weeded out, and trustworthiness scores would accumulate.
Trustworthiness ratings on pages editable only by experts would relieve the need for internal citations. Right now, the Wikipedia community pushes hard for citations in articles. Citations are important in the Wikipedia context because of the lack of credentials on the part of the writers. But if pages were only editable by authenticated lawyers, then cumulative positive ratings would make pages more reliable even without citations.
Admittedly, a forked version of Wikipedia edited by lawyers and law students would not replace the big binder sets. The depth of the material, at least as it stands now, is too limited. Wikipedia, even if reliable, wouldn't help a trust-and-estates lawyer with trust and estates. But if it were imbued with trustworthiness, Wikipedia content does have enough depth to be useful for lawyers orienting themselves to an unfamiliar field of law. Likewise, it has enough detail for non-lawyers who are looking to gain a general understanding of some specific doctrinal topic.
So, what do you think? It would be fairly easy to put together from a technical perspective, but would it be worthwhile? Do you think people with legal knowledge would contribute by removing errors and scoring pages? Or would a forked law wiki sit fallow? (I do notice there are lots of lawyers participating actively on Quora, crafting very good answers to individual questions.)
Maybe it would work in other fields aside from law, as well. It does seem to me, though, that law is a particularly good subject for forking.
TrackBack URL for this entry:
Listed below are links to weblogs that reference A Plan for Forking Wikipedia to Provide a Reliable Secondary Source on Law:
I've actually found Wikipedia's entries on legal topics to be impressively accurate on the whole. Definitely would be easier to take on faith with something like this though.
Posted by: Patrick | Jul 29, 2011 7:48:19 PM
You are aware of Citizendium, right?
Posted by: James Grimmelmann | Jul 30, 2011 12:37:15 AM
You could also simply vet particular (static) revisions of an article on Wikipedia itself. Wikipedia supports this already to some degree; see for example [[Talk:Conatus]].
Speaking as a Wikipedian, I'd love to see an expansion of that sort of participation, because it allows for a bit of both the mass participation and the focused, expert review—some of the best of both worlds.
If you'd like to organize something along those lines, I can help! :)
Posted by: Nihiltres | Jul 30, 2011 1:06:32 PM
And maybe it's too early to call it, but the fact that Citizendium currently only has 156 "approved" articles across all subjects goes a long way to answering some of the questions posed here.
I also think it's wrong to assume that Wikipedia's legal articles aren't trustworthy, credible, etc. I've never seen a study focused on the quality of Wikipedia's legal coverage, but when Wikipedia has been put to the test against mainstream encyclopedias in other areas, it's always held its own fairly well.
If you want to improve Wikipedia, that's what the edit button is for.
Posted by: Brian | Sep 14, 2011 1:47:49 AM
The comments to this entry are closed.